[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
Re: Sky high load average
- From: Aaron Konstam <akonstam Trinity Edu>
- To: enigma-list redhat com
- Subject: Re: Sky high load average
- Date: Sat, 16 Nov 2002 20:33:03 -0600
On Sat, Nov 16, 2002 at 08:37:22AM -0800, Mark Denni wrote:
> Aaron -
>
> Can you explain in more detail. I don't know how to
> check context switching. Were you able to fix the problem?
> Would renicing the jobs as they are starrted help?
> Is there some configuration I need to change, or some
> patch I need to apply?
>
> - Mark
>
The program vmstat shows context switches per second as cs in the table it
produces. That is only a symptom; what to do about it is not so obvious.
In our case it was a server and we think that it was processing the ext33
journal files that were causing the problem. Our load averages were high
but the cpu usage was not that high.
Now I am not against the solutions others are providing. NFS from a large
number of machines can also cause high context switching. I just mention
this because by monitoring the context switching as you make changes in the
system will be another way to see if you are getting anywhere toward a
solution or if high context switching is even your problem.
--
-------------------------------------------
Aaron Konstam
Computer Science
Trinity University
715 Stadium Dr.
San Antonio, TX 78212-7200
telephone: (210)-999-7484
email:akonstam trinity edu
[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]